What is LINEAR PHASE eq?

[logancircle]

Real or wives-tale? I've heard good things about them, like they avoid creating artifacts that cheaper, grainy EQs don't. Like if you make a dip in the 200Hz, a non-linear phase EQ creates weird harmonic junk several octaves above. ?

[Mystic Steamship Co.]

It's an eq that doesn't cause any (or maybe just not as much) phase shifting as a normal eq. They sound pretty wierd to me, but the only one i've used is the waves plug in.

[calaverasgrandes]

my recollection is a little muddy. I'll blame it on print through. But I seem to remember that there are three gotchas of eq circuits. phase, group delay and noise. The gotcha being that when you go more in the direction of reducing one, the others will increase. So a linear phase eq would have problems with group delay or noise.
Caveat emptor, its been a while since I even attempted to design a circuit so I am pretty rusty on this area. Check out Traister's books on power supply design and diodes for more info.
But back to the point, an analogue eq with "perfect" phase behaviour, will have a shortcoming in some other area. BTW a grainy EQ isnt such because of phase. More likely its distortion and/or noise. Obnoxiously phasey EQ usually sounds plasticy and smeary to my ears. Like most Ramsa and Toa mixers have EQ that I dont like.

[The Scum]

Yes, it exists. It's somewhat a marketing-hype term that corresponds to a particular type of fundamental design.

Here's the most concise description I can find:

http://en.wikipedia.org/wiki/Filter_des ... al_filters

You'll notice in the IIR description, they mention that the frequency response is actually determined by the phase shift. If you "slow down" 200 Hz, it'll start to cancel itself out of the original signal...conversely if you "speed it up," it'll reinforce itself.

FIR = linear phase EQ.

[somehow, my brain always likens this to a frequency-dependent doppler effect of sorts, but I'd have to draw a bunch of pictures to describe it any better than that.]

There are also some techniques in analog to implement linear-phase EQ by adding a bunch of circuitry to counter the phase shift of the EQ filter. I think the Amek 9098 stuff did that...

I also wonder if what a lot of people call "phase" in reference to an EQ is really related to the necessary phase-shift in the design, or if it's an artifact of the sound. I think some of it comes from the fact that when you sweep a very narrow notch or boost (very high Q), the result sounds like a phase shifter pedal...you aren't so much "hearing the phase" as you are the damaging artifacts of a pretty severe filter.

[vibesof20hz]

I have 2 cents in my pocket!

Normal band pass EQs work by shifting the phase relationship of one portion of the original signal against itself. This of course generates artifacts in the waveform at frequencies other than the center frequency, and its surrounding frequencies. However, while cutting 4dB at 100Hz with a narrow bandwidth (which can be calculated by Log 10 (frequency 1 / frequency 2) / 0.301, in case you wanted to do that) you would find that only a narrow frequencies are effected by the eq.

However, because of the nature of eq that effects phase, the effected frequencies become the fundamental frequency for a whole series of harmonics which can shape the timber of the whole signal. As greater boosts and cuts are applied, some frequencies are time-shifted more than others, essentially pulling apart the precise alignment of overtones that define a musical instrument's character.

Now, many people may consider this a bad thing. Most of the time when I touch a Behringer EQ I have this intense urge to punch myself in the face due to the effects of this phenomena (among other things). However, when I listen through a Pultec, the same phenomena is going on, except this time, it sounds good to most people. So, while this ?phase error? or whatever you may want to call it is generally considered a bad thing, it also may be a good thing, depending on who you talk to.

Linear phase EQs function the same way, however, the shift is the same for all frequencies in the band pass. This avoids the harmonics induced by the ?phase shift? phenomena. Mostly, the linear phase EQs are used for mastering for reasons that go beyond a post (refer to Bob Katz's books for starters), but now that you can download a plugin for free, its al up for grabs.

[river]

I bought one of the earliest phase linear EQ plugins, the PAR PLEQ3. It's definitely best used for mastering, being a major CPU vampire. Still, when a mix has undesireable phase buildup that creates that gauzey veil, merely putting the mix through this EQ without even cutting or boosting any frequencies is like Windex on a dirty mirror. It brings clarity and better stereo imaging to a muddy mix. It's not right for every mix, sometimes you want the smear.

[skyguy]

Most of the time when I touch a Behringer EQ I have this intense urge to punch myself in the face due to the effects of this phenomena

[logancircle]

Maybe we could punch him in the face and take the Pultec... Just an idea. Not a good one, tho.

[thunderboy]

I love my IRP Transversal EQs!

jt

[calaverasgrandes]

All I got to say is, once I read about how EQs actually work in the analog domain, I found I couldnt bear to touch an eq at all for a long time. then I thought I was off the hook with digital eqs. Not so much. I dont know the nitty gritty, but with most of my eq plug ins, the cumulative result is lacking. the only EQ I dont get sick of using is the TC EQsat. I dont know if its supposed to be a "tube" sound eq or what. It just does a nice highpass without butchering the low mids. It also can boost highs without making it sound like it was all recorded through a behringer. Too bad I dont have enough powercore "power" to use it on all my tracks all the time.

[Catfish]

What was that motto attributed to Joe Meek????????







Catfish...........

[leigh]

I bought one of the earliest phase linear EQ plugins, the PAR PLEQ3. It's definitely best used for mastering, being a major CPU vampire. Still, when a mix has undesireable phase buildup that creates that gauzey veil, merely putting the mix through this EQ without even cutting or boosting any frequencies is like Windex on a dirty mirror. It brings clarity and better stereo imaging to a muddy mix. It's not right for every mix, sometimes you want the smear.

So, when you have all the controls set flat, it's still doing something? Not what I'd expect from a "surgical" tool like linear-phase EQ.

To be clear - are you talking about this Algorithmix EQ, or something else?

[river]

So, when you have all the controls set flat, it's still doing something?

Sounds unlikely, but yes, it is affecting the mix. It's necessary to activate the 3 bands, but not necessary to cut or boost any frequencies (although I often do while mastering a mix). I got the name of the plugin slightly screwed up, it's the PL ParEQ3 by Refined Audiometrics.

[leigh]

Sounds unlikely, but yes, it is affecting the mix. It's necessary to activate the 3 bands, but not necessary to cut or boost any frequencies (although I often do while mastering a mix). I got the name of the plugin slightly screwed up, it's the PL ParEQ3 by Refined Audiometrics.

Thanks for the clarification. I'm curious as to what's going on with this plug, but since it's PC VST only, I won't be finding out soon.

They offer a free one-band version of their linear-phase EQ, for those interested...